Deep Learning Captures More Accurate Diffusion Fiber Orientations Distributions than Constrained Spherical Deconvolution

TL;DR

This paper demonstrates that deep learning can extract more accurate intra-voxel fiber orientation distributions from diffusion MRI than traditional constrained spherical deconvolution, revealing additional information in the diffusion signal.

Contribution

The study introduces a deep learning model that surpasses CSD in capturing fiber orientations from single shell diffusion MRI data, highlighting new information in the diffusion signal.

Findings

Deep learning outperforms CSD in fiber orientation estimation.

Additional information exists in diffusion signals beyond CSD's current use.

The model aligns well with histological fiber distributions.

Abstract

Confocal histology provides an opportunity to establish intra-voxel fiber orientation distributions that can be used to quantitatively assess the biological relevance of diffusion weighted MRI models, e.g., constrained spherical deconvolution (CSD). Here, we apply deep learning to investigate the potential of single shell diffusion weighted MRI to explain histologically observed fiber orientation distributions (FOD) and compare the derived deep learning model with a leading CSD approach. This study (1) demonstrates that there exists additional information in the diffusion signal that is not currently exploited by CSD, and (2) provides an illustrative data-driven model that makes use of this information.